Compose is a machine learning tool for automated prediction engineering. It allows you to structure prediction problems and generate labels for supervised learning. An end user defines an outcome of interest by writing a labeling function, then runs a search to automatically extract training examples from historical data. Its result is then provided to Featuretools for automated feature engineering and subsequently to EvalML for automated machine learning. Prediction problems are structured by using a label maker and a labeling function. The label maker automatically extracts data along the time index to generate labels. The process starts by setting the first cutoff time after the minimum amount of data. Then subsequent cutoff times are spaced apart using gaps. Starting from each cutoff time, a window determines the amount of data, also referred to as a data slice, to pass into a labeling function.
Features
- By automating the early stage of the machine learning pipeline, our end user can easily define a task and solve it
- Machine learning tool for automated prediction engineering
- It allows you to structure prediction problems and generate labels for supervised learning
- An end user defines an outcome of interest by writing a labeling function
- Its result is then provided to Featuretools for automated feature engineering and subsequently to EvalML for automated machine learning
- The label maker automatically extracts data along the time index to generate labels
Project Samples
